-
An Introduction to Sediment Transport in Estuaries
Larry Sanford [email protected]
Gail [email protected]
-
Outline
Introduction to basic principles of sediment transport Emphasis
on erodibility of muds
Modes of transport Modeling High Concentration Suspensions
Turbidity Maxima
-
General References for figures, etc. (not always noted)
Allen, J.R.L., 1985. Principles of Physical Sedimentology.
George Allen & Unwin Ltd., London, Boston, Sydney, 272 pp.
Madsen, D.O.S. and Wood, D.W., 2002. Sediment Transport Outside
the Surf Zone. In: D.T. Walton (Editor), Coastal Engineering Manual
Outline, Part III, Coastal Sediment Processes, Chapter III-6,
Engineer Manual 1110-2-1100. U. S. Army Corps. of Engineers,
Washington, D. C., pp. 72.
Mehta, A.J. and McAnally, W.H., in press. Cohesive sediments,
Sedimentation Engineering, Manual 110. ASCE.
Open-University, 1999. Waves, Tides, and Shallow Water
Processes, Second Edition. The Open University, Milton Keynes, UK,
227 pp.
van Rijn, L.C., 1993. Principles of Sediment Transport in
Rivers, Estuaries and Coastal Seas. Aqua Publications, 386 pp.
Winterwerp, J.C. and Van Kesteren, W.G.M., 2004. Introduction to
the Physics of Cohesive Sediment in the Marine Environment, 56.
Elsevier B.V., Amsterdam, The Netherlands, 466 pp.
Wright, L.D., 1995. Morphodynamics of Inner Continental Shelves.
CRC Press, Boca Raton, 241 pp.
-
Basic Characteristics of Sediments
They sink or settle (Latin sedere, to sit) and have inertia They
dont move exactly with the water
There are sources (rivers, shorelines, atmosphere, bottom) and
sinks (bottom)
Particle behavior depends on size, weight, stickiness, and
shape
Biology can affect physics at lowest order Other particles
(i.e., plankton) can be thought of
as sediments with different density and behavior
-
Global Sediment Sources Rivers account for 85% of inputs to
global ocean
-
Estuaries can trap large portions of riverine inputs, can have
large internal inputs from shoreline erosion, and can import
sediment from the ocean (from Langland and Cronin 2002)
-
The bottom can serve as both source and sink, and often
dominates both terms in estuaries and coastal seas
4,400 MT/d
90,000 MT resuspended
+ 45,000 MTbackground
=135,000 MT
0 MT/d
184,400 MT/d180,000 MT/d
4,400 MT/d
Upper Chesapeake Bay Suspended Sediment Mass Balance
-
What matters the most for sediment transport dynamics? Not
minerology, stickiness, or size per se, but rather settling speed,
remobilization criteria, and deposition criteria.
-
Background Classification of Sediment Transport (in decreasing
order of understanding)
Non-cohesive > 64 um particle size, including coarse silts,
sands, and gravels,
interparticle cohesive forces negligible, highly permeable
Dominant on energetic inner continental shelves far from sources
of
fines Previously thought to be biogeochemically boring, under
revision
Cohesive < 64 um particle size, poorly sorted mix of silts,
clays, and organics
(mud), interparticle cohesive forces dominate, impermeable
Dominant in less energetic environments and/or close to sources
Strong correlation to transport and fate of POC and associated
contaminants Primary determinant of turbidity
Mixed Sands with > 10% mud, muds with > 5-10% sands,
essentially
impermeable Dominant in all other environments More resistant to
erosion than either sand or mud alone (?) Very complex and poorly
understood dynamics
-
What determines the settling velocity of a sediment
particle?
-
Stokes settling velocity
-
Drag coefficient around a sphere (from ?, one of the general
refs, I think)
-
Alternative approximation due to Gibbs (1971) as quoted by
Wright (1995)
Settling velocity for higher Re (larger or more dense)
1/ 22 23 9 1 (0.003869 0.02480 )
0.011607 0.07440
s
s
gD Dw
D
+ + + = +
-
Now, what factors affect remobilization of sediment from the
bottom?
Trade-off between forcing (bottom shear stress) and response
(resistance to motion)
Bottom shear stress important for estuarine sediment transport
is some combination of steady stress (tidal and longer period
forcing) and high frequency oscillatory stress (surface waves)
Except for completely smooth bottoms, stress must also be
separated into the total stress that affects the flow above the
bottom, and the skin friction that affects the sediment bed
Resistance to motion and transport behavior can be quite
different for non-cohesive and cohesive sediments, and in general
is much more complex for cohesive sediments
-
Separation of total stress and skin friction for quasi-steady
flows(not the most sophisticated technique, but has been used
previously [Glenn and
Grant, 1987; Grant and Madsen, 1982] and is simple)
* 11
0( ) lnT
T
u zu z z
=
2 21* ( )T T du C u z = =
2
1
0
lnd
T
C zz
=
Assume we have a reasonable estimate of the total stress bottom
roughness parameter z0T from model calibration or direct stress
measurements in the log layer. Then the log layer velocity law (as
ably explained by Steve Monismith) gives
where u(z1) is the velocity at height z1, u*T is the total
stress shear velocity, and =0.4 is von Karmans constant. This can
be solved for the total hydrodynamic bottom stress as
where
-
Now write a new law of the wall for the skin friction shear
velocity u*s in terms of u(z1) determined above and the skin
friction roughness parameter z0s
1 1
0 0* 1 * 1
1 1 1
0 0 0
ln ln( ) ( )
ln ln lnT T
s T d
s s s
z zz zu u z u C u zz z z
z z z
= = =
And the ratio of the skin friction s to the total stress T
is
21
0
1
0
ln
lns T
T
s
zzzz
=
Ratio of skin friction to total stress for kb,tot = 3 cm,
kb,skin=0.3 mm, 10
sigma layers
00.20.40.60.8
1
0 5 10 15 20Water Depth (m)
S
k
i
n
f
r
i
c
t
i
o
n
/
T
o
t
a
l
S
t
r
e
s
s
-
Surface gravity waves (only most important aspects for sediment
transport)
-
bau ghh
=
0bu =
is the amplitude of the near bottom velocity fluctuations
bu
-
From Wright (1995)
-
As a deep water wave moves into shallower water, it first
becomes a transitional wave and eventually a shallow water wave.
Deep is defined relative to wavelength (or period), however, and it
is different in different environments. Wave base refers to the
depth at which typical waves in that environment first begin to
significantly influence the bottom.
-
Oscillatory boundary layers and wave friction factors
* /w wu 21
2bw w bf u =
Oscillatory boundary layers have an inherent vertical length
scale associated with the diffusion limit for turbulence in one
cycle,
Wave bbls are typically cms thick, which greatly enhances shear
and turbulent stress
This leads to much larger drag coefficients fw
* /w wu
When compared to quasi-steady drag coefficients Cd
-
So surface gravity wave forcing can play a dominant role in
sediment transport in shallow, microtidal environments
(Nakagawa, Sanford, and Halka 2000)
Simulated significant wave height (m) Estimated bottom shear
stress (Pa)
Simulated wave height and wave induced bottom shear stress for a
15 m s-1wind from the northwest in Baltimore Harbor, Maryland,
USA
-
Figure 3 Time series observations of mud resuspension before,
during and after a storm, from a site in upper Chesapeake Bay,
Maryland, USA. (Sanford in press)
-
Resistance to remobilization of bottom sediments:Critical
Stress
A threshold value of the applied skin friction below which there
is no (or negligible) transport of bottom
sedimentsFor non-cohesive sediments, the stabilizing force Fg is
the submerged weight of the particle
-
Sh ields parameter / determines whether a particle will move.
For anygiven particle (s, D), there is a critical value of above
which sed imentmotion occurs called the critical Shields parameter
6 c. c often plotted as a function o f (Re* =u*kb/). For a flat
sediment bed,kb - D50 / med ian diameter o f the sediment grains,
so
The value of c mu st be empirically determined, resultin g in
ShieldsDiagram:
c often plotted as a function o f (Re* =u*kb/). For a flat
sediment bed,kb - D50 / med ian diameter o f the sediment grains,
so
The value of c mu st be empirically determined, resultin g in
ShieldsDiagram:
50*c
u D= func ( )
-
**
(c50
50
= funct S )Dwhere S = g(s - 1)D4
This is inconvenient because both axes are a function of u*, so
derive a modified Shields diagram (Madsen and Grant 1976); also
valid for waves
-
A simpler, but much less general way to present this information
is to assume quartz grains with a fixed drag coefficient, and
derive a relationship for motion in terms of grain size and current
speed (at right). This is sometimes referred to as a
Hjulstromdiagram.
This diagram also indicates a fixed erosion threshold for silts
and clays, which is overly simplistic, and a general behavior known
as exclusive deposition, which is not universally accepted.
From Waves, Tides, and Shallow Water Processes (1999)
-
Erosion/Resuspension of silt/clay mixtures (muds)
Resistance of the bed a function of cohesion, water content,
grain size distribution and density. Often expressed as
erodibility, parameterized by critical stress c and erosion rate
E.
In general, mud erodibility is not predictable a priori and at
least some measurements are required
-
Partial list of factors affecting fine sediment erodibility
Factor Sense of
effectWater content +Percent sand -
Exopolymers (EPS)/Mucopolysaccharides
-
Bioturbation +
Percent clay -
Salinity -
From Roberts et al. (1998)
-
Erodibility also can vary significantly in time and space. For
example, consolidation causes c to increase rapidly with
depth into the bed and with time after deposition (Parchure and
Mehta 1985)
-
Erodibility can change significantly in response to
disturbance.Passage of a tropical storm, upper Chesapeake Bay, Sept
1992.
Dredged sediment disposal site, 5 m depth.
Stn 6, 1m
Stn 8, 3m
Stn 6, 1m
Howell Pt. Buoy
September 1992
17 18 19 20 21 22 23 24 25 26 27 28 29
S
p
e
e
d
(
c
m
/
s
)
0102030405060
R
M
S
P
(
m
b
)
02468
10
T
S
P
(
m
g
/
l
)
050
100150200
(
W
i
n
d
S
p
e
e
d
)
2
050
100150200250
-
10 cm
Physical disturbance overwhelms bioturbation
at York River site
Bioturbation dominates fine sediment fabric at Chesapeake Bay
site
Interactions between physical and biological disturbance of the
sea bed can lead to distinct layering or near homogeneity
( x-radiographs courtesy of Linda Schaffner)
Increasing physical disturbance of sea bed
-
Peter TraykovskiHigh Resolution Acoustic Backscatter Profiler,
Hudson ETM
Present state of the sediment bed can reflect a very dynamic
interaction between rapidly repeated erosion and deposition
-
Recent emphasis in field has been on development of new
techniques for site-
specific, in-situ erosion testing
Laboratory tests on sediment samples from field used to be
common, now less frequent
In situ bottom landing devices Annular and linear flumes
Small-scale erosion chambers
Quasi in situ testing of cores Flumes Small scale erosion
chambers
-
In situ annular and linear flumes
VIMS Sea Carousel Jerome Maa (similarto Carl Amos device at
SOC)
NIWA Linear Flume (Aberle et al. 2004)- Similar to Tom Ravens
device in US
-
In situ small scale erosion chambers
10 cm diameter SedErode, successor to ISIS, Williamsonand
Ockenden (1996)
Cohesive Strength Meter, sold by Partrac Ltd, UK
-
Quasi in situ core testing
UMCES 10 cm diameter MicrocosmSedFlume, McNeil et al. (1996)
-
The UMCES Microcosm Core Erosion System
With thanks to Drs. Giselher Gust and Volker Mueller, TUHH
-
Huge scatter in reported erosion rates may be due to several
factors:
Real differences in erodibility what we want to understand and
predict
Differences in instrument calibration or performance what we
need to know in order to compare different data sets
Differences in experimental design and data analysis what we
need to resolve before we can compare instruments
From Gust and Morris (1989)
-
How do experimental design and data analysis procedures affect
interpretation of erosion data? For illustration,
consider data from an erosion device intercomparisonexperiment
in Upper Chesapeake Bay, May 7, 2002
VIMS Sea Carousel UMCES MicrocosmFig. 1-1. VIMS Sea Carousel
Experiment Sites for the Middle Chesapeake Bay (M1 and M2) and
Upper Bay Sites (U1 and U2).
-
We have known for years that a rapid increase of c with depth
into the bed results in a rapidly time varying erosion rate:
Example erosion test and derived Example erosion test and
derived critical stress profile from Parchure and critical stress
profile from Parchure and
Mehta (1985)Mehta (1985)
-
00.05
0.1
0.15
0.2
0.25
0 5000 10000 15000Time (s)
E
r
o
d
e
d
M
a
s
s
(
k
g
m
-
2
)
0
0.2
0.4
0.6
S
h
e
a
r
S
t
r
e
s
s
(
P
a
)
observed Shear
If erosion behavior is time dependent, then differences in the
time history of stress application
affect the results of erosion experiments
Most common to apply a sequence of constant stress steps,
ranging from minutes to hours in duration
Sometimes gradually ramp up stress between stress steps (Maa et
al. 1993)
Others apply continuously increasing stress (Gust and Morris,
1989)
Others apply short bursts of stress of increasing intensity (CSM
of Tolhurst et al., 2000)
1.E-06
1.E-05
1.E-04
0 3000 6000 9000 12000 15000Elapsed time (s)
E
r
o
s
i
o
n
R
a
t
e
(
k
g
s
-
1
m
-
2
)
0
0.2
0.4
0.6
S
h
e
a
r
S
t
r
e
s
s
(
P
a
)
GP1 Shear
1.E-06
1.E-05
1.E-04
0 3000 6000 9000 12000 15000Elapsed time (s)
E
r
o
s
i
o
n
R
a
t
e
(
k
g
s
-
1
m
-
2
)
0
0.2
0.4
0.6
S
h
e
a
r
S
t
r
e
s
s
(
P
a
)
GP1 Shear
Microcosm
0
0.05
0.1
0.15
0.2
0.25
0 5000 10000 15000Time (s)
E
r
o
d
e
d
M
a
s
s
(
k
g
m
-
2
)
0
0.2
0.4
0.6
S
h
e
a
r
S
t
r
e
s
s
(
P
a
)
observed Shear
Sea Carousel
-
An unexpected time dependent erosion response leads to
differences in interpretation or reporting of
erosion rate data
Many interpret erosion rate for each applied stress as the
average over the time interval
If erosion rate is time dependent, then the average depends on
the duration of the interval
Others interpret erosion rate as the average after the initial
spike in response (Ravens and Gschwend 1999)
Others report erosion rate as the initial spike (Maa et al.
1998)
Others interpret erosion rate as an expected time varying
response to changing erodibilityduring the interval (Sanford and
Maa 2001)
These differences in interpretation contribute to the orders of
magnitude differences in erosion rates observed, even in controlled
inter-comparison experiments (Tolhurst et al. 2000)
From Aberle et al. (2004)
-
Differences in interpretation of critical stress are also
associated with time/depth dependence
Many investigators focus on a single critical stress for
erosion, the value at which initiation of motion occurs Some define
critical stress as that stress for which a significant
increase in concentration is first observed Others define
critical stress as the zero intercept of an erosion rate
vs stress regression Others define critical stress as a depth
profile of stresses
at which erosion stops Much more common in older laboratory
erosion tests, seldom
reported for in situ measurements Others consider the concept of
critical stress to be
unimportant, or define it but do not use it in reporting erosion
rate results
-
Sea Carousel Microcosm
teEE = 0
globally )( 20 cbE
0
0.05
0.1
0.15
0.2
0.25
0 0.1 0.2 0.3 0.4c (Pa)
E
r
o
d
e
d
m
a
s
s
(
k
g
m
-
2
)
00.010.020.030.040.050.060.070.08
0 0.0002 0.0004 0.0006 0.0008 0.001
M (kg s-1 Pa-1 m-2)
E
r
o
d
e
d
m
a
s
s
(
k
g
m
-
2
)
)]()()[( mtmME cb =
locally )( cbE
ncbE )(0
Sea Carousel v. Microcosm comparison using individual data
analysis and interpretation techniques
-
0for 0,
' ( )
b
tb c
ddt
E e
=
=
assume constant,
then for n=2 and 0,
b
b
ddz
ddt
=
=
00, where 1
bE dE B E kBt dz
= =+
Time dependent linear 2 parameters,c depends on depth of erosion
(Sanford and Maa 2001)
Time dependent power law 4 parameters, simple extension of
Roberts et al. (1998)
Time dependent erosion behavior is relatively straightforward
toderive theoretically, especially for specific erosion
formulations
that yield analytical solutions:
)()( zknb etAE =
Assume constant, cd zdz = =
-
Reanalysis of Sea Carousel data using the Microcosm approach
reveals real similarities
and differences
0
0.05
0.1
0.15
0.2
0.25
0 5000 10000 15000Time (s)
E
r
o
d
e
d
M
a
s
s
(
k
g
m
-
2
)
0
0.2
0.4
0.6
S
h
e
a
r
S
t
r
e
s
s
(
P
a
)
observed fit Shear
1.E-06
1.E-05
1.E-04
2450 4450 6450 8450Elapsed time (s)
E
r
o
s
i
o
n
R
a
t
e
(
k
g
s
-
1
m
-
2
)
0
0.2
0.4
0.6
S
h
e
a
r
S
t
r
e
s
s
(
P
a
)
GP1 fit Shear
0.00
0.05
0.10
0.15
0.20
0.25
0.1 0.15 0.2 0.25 0.3 0.35 0.4Critical Stress (Pa)
E
r
o
d
e
d
m
a
s
s
(
k
g
m
-
2
)
Sea CarouselMicrocosm
0.00
0.05
0.10
0.15
0.20
0.25
0 0.0005 0.001 0.0015 0.002 0.0025Erosion Rate Constant (kg s-1
Pa-1 m-2)
E
r
o
d
e
d
m
a
s
s
(
k
g
m
-
2
)
Sea CarouselMicrocosm
Microcosm
Sea Carousel
-
Measurement Conclusions Ultimate goal is to predict fine
sediment erosion
behavior with a minimum of site-specific data Reliable data
comparisons hindered by
differences in instruments, experimental designs, and data
interpretation/analysis
Standardization of experimental design and data analysis
protocols needed Should be informed by appropriate choice of
erosion
model(s) Accounting for time/depth changes in erodibility
especially important time dependent erosion response should be
expected, not ignored
-
Measurement Conclusions, continued
Experimentalists should archive and share erosion test time
series, not just derived parameters
Erosion experiments should be considered as spot measurements of
a dynamically varying process
Modelers should match formulations in sediment transport models
to formulations used for data collection/analysis
-
DepositionDeposition
-
Time
e b
TSL
d
Exclusive, unlimited sediment supply
Exclusive, limited sediment supply
Continuous, unlimited sediment supply
-
From Sanford and Halka (1993)
-
Changeover
-
Modeling Cohesive Sediments: Erosion, Deposition, and Bed
Processes
-
What are the main challenges for fine sediment transport
research and application? Effective shear stress Sediment
erosion/resuspension Shoreline erosion/protection Spatial
heterogeneity Flocculation and settling Sediment Deposition and
Consolidation Biological effects Concentrated benthic suspensions
Interactions with T/S stratification Large events Modeling
Effective adoption of new technologies Disciplinary and
geographical boundaries and biases
-
Bottom Sediment Erosion/Resuspension
Resistance of the bed a function of cohesion, water content,
grain size distribution and density. Often expressed as
erodibility, parameterized by critical stress c and erosion rate
E.
In general, erodibility is not predictable a priori and at least
some measurements are required (a sad state of affairs)
Variety of methods to test erodibility exist, but no real
standards for collection, interpretation, or use of the data
Can measure, model erodibility profile at one point in time, but
how does the bed evolve from that point or recover from
disturbance? Consolidation Armoring Bioturbation/bio-adhesion
When to switch between fluid mud entrainment and cohesive bed
erosion?
Mass erosion bed failure under high stresses, rather than
particle by particle
-
Biological effects
Particle stickiness, repackaging, biodeposition Bottom roughness
of benthic communities Bioturbation mixing of surface sediments
changes surface texture, erodibility Macroscale structure/form
drag - potential for
feedbacks to sedimentation Oyster beds and reefs, coral reefs
Seagrass beds Marsh vegetation
-
Concentrated benthic suspensions
Some major advances in recent years AMASSEDS, STRATAFORM,
COSINUS programs Turbulence damping at lutoclines Turbidity
flows
CBS are common where energy is high and sources are large, but
just a slight decrease in energy can result in a change to normal
sediment beds Is this a state change? Why, and when is the
switch
thrown?
-
Spatial heterogeneity
Changes in bottom sediment texture and strength can be abrupt,
sometimes at smaller scale than model cells Why? Is this a state
change, rather than
gradual mixing? Can we model it? Recent work of van Ledden et
al. provides a
framework for interpretation/modeling New measurements are
fantastic, but how
do we utilize the information to improve understanding and
prediction?
-
Modeling How to best incorporate all of the above into
workable models? 1-D process models are important, especially
for
development, but 3-D long term, large scale models are
critically needed
Detailed knowledge of processes must be parameterized, but
retain essential features
Best models are as simple as possible, but no simpler
The NOPP Community Sediment Transport project a major
opportunity
What can be adopted from previous work, and what needs new
development?
-
NOPP CSTM participants and their roles
-
Selected cohesive sediment erosion formulations
Gularte et al. (1980), others
Lick (1982), many others
McLean (1985), many others
Sanford and Maa (2001)
)()( zknb etAE = Roberts et al. (1998), Lick et al. (2006)
I
II
III
IV
V
-
Consolidation effects on erosion
Expressions I, III, and V explicitly allow for a consolidation
effect
Lick and co-workers have modified II to allow for a
consolidation effect according to
Keen and Furukawa (2006) also modified II to allow for a
consolidation effect (as well as bioturbation) using multiplicative
factors parameterized based on observations
However, few sediment transport models include both
time-dependent consolidation AND consolidation effects on erosion
(time-dependent, supply-limited erosion)
-
Time dependent erosion behavior is straightforward to derive for
specific erosion formulations :
dStart with ( )[ ( ) ( )], take , and set dt
cb c
dE M m t mdm = =
get bddE ME Mdt dt
+ =
if , then andbbdAt Adt = =
( ) ( )0 01 Mt Mtb cAE e M e = + 1set , average over time step
and set
teM t bt
= = ( ) ( )0 0then 1 b cAE b bM = +
Time dependent behavior is important when b 1).
-
What would an ideal cohesive bed/erosion/deposition model
include?
Physically correct consolidation model (based on modified Gibson
equations and including both phases of consolidation) including
sand-mud mixtures, erosion/deposition, discontinuities
General representation of erosion behavior Including arbitrary
sand-mud mixtures Including dependence on consolidation state Time-
and depth-dependent erosion behavior (limited sediment
supply as a function of applied stress) Parameterization of
changes in erosion behavior between fluid
mud entrainment, surface erosion, and mass erosion
Parameterizations for bioturbation and bio-adhesion
(both erosion behavior and bed evolution) Deposition formulation
that allows for multiple particle
classes, floc behavior, and high erodibility of recent
deposits
Compatibility with non-cohesive model components
-
Selected Existing Cohesive Sediment Transport Models
SEDZLJ (Jones and Lick 2001) includes sediment mixtures with
empirical cohesive erosion
parameterizations, but limited bed dynamics Now being
incorporated into ECOMSED (Hydroqual) with Sanfords
consolidation mechanism for deposited sediments DELFT3D
Proprietary code, but containing much of the high quality work
done at TU Delft (Winterwerp, Kranenberg, van Kesteren, and their
students and post-docs)
Rumored to become open source soon SEDTRANS05 (Neumeier et al.
2007)
Latest incarnation includes Gularte type cohesive erosion with a
limited empirical approach to equilibrium consolidation
capability
EFDC (Hamrick, Tetratech) Includes sophisticated, but complex
algorithms to handle mixed
sediments and consolidation (I think), but not easily accessible
OTHERS (MORPHOS, WES CH3DSED, OSU CH3DSed(?), ?) ROMS-SED Warner et
al. in press, under development
-
Sediment Bed Model Development in ROMSSED based on Sanford (in
press)
Use a layered bed model with continuous profiles of c,
layer-averaged erosion constant M and sand fraction fs
Use sediment bed mass m as independent variable instead of
depth(better for consolidation)
2-component mixture of sand and mud Separate erosion parameters
for sand and mud (interaction effects
not yet incorporated) Erosion rates proportional to fractions at
interface
Mud erosion follows Sanford and Maa (2001) Sand erosion follows
Harris and Wiberg (2001) Assume constant c,sand Assume that c,mud
approaches an equilibrium profile at a first order rate Allow for
sediment mixing (bioturbation, bedload transport) Assume that newly
deposited mud particles carry with them a very low c,mud, which
lowers the critical stress of the sediment surface layer Interface
moves down/up through bed layers during erosion/deposition Only mix
mass between layers when a threshold is exceeded (minimizes
numerical dispersion) Model evolution of c, mud and fs as a
function of m and time
-
How well does an exponential approach to equilibrium approximate
consolidation?
Comparison to data of Toorman and Berlamont (1993)
Original data and full consolidation model by T&B (1993)
T&B (1993) data, exponential approach to equilibrium
approximation
-
Example: Erosion, deposition, and consolidation of a pure mud
and a sand-mud mixture
Bed consists of 25 layers 0.05 kg m-2 thick Critical stress
profile initiated with average of Baltimore
Harbor profiles from Sanford and Maa (2001), also assumed to be
equilibrium profile
Assume M=ss where = 11.75 m d-1 Pa-1 from SM2001
Spring-Neap cycle of tidal shear stress, max varies between
0.15-0.30 Pa
A 1.25-day event starting at day 21.25 increases the max stress
2.5X
Very low sediment mixing of 0.01 cm2 yr-1 wsm = 86.4 m d-1, h =
2 m Consolidation rate = 3.0 d-1, swelling rate = 0.03 d-1 wss =
864 m d-1, csand = 0.125 Pa (fine sand)
-
0 5 10 15 20 250
0.1
0.2
0.3
0.4
T
S
S
[
g
.
l
-
1
]
0 5 10 15 20 250
0.30.60.9
t
a
u
b
[
P
a
]
m
[
k
g
.
m
-
2
]
0 5 10 15 20 25
0
0.5
1
m
[
k
g
.
m
-
2
]
time, [d]
0 5 10 15 20 25
0
0.5
1 00.20.40.6
All mud, very low sediment mixing
-
All mud, very low sediment mixing, 2 days during event and 1
tidal cycle 2 days after event
-
30/70 sand-mud, very low sediment mixing
-
30/70 sand-mud, sediment mixing 10 cm2 yr-1
-
Modeling Conclusions Layered bed model for critical stress
profile in
terms of bed mass simplifies formulation, avoids layer transfers
during consolidation. Z-dependent model also possible
Specification of equilibrium conditions based on observed
erosion behavior promising, but may need tweaking.
Consolidation formulation predicts reasonable behavior with
little computational effort, but needs more validation
Mud-sand mixture and diffusive mixing schemes lead to realistic
complex bed structures and directly affect resuspension
-
Modeling Conclusions, continued
Need to incorporate sand-mud interaction effects on
erodibility
Need to simplify code, translate into Fortran for inclusion in
ROMSSED (in progress)
-
Changeover
-
Estuarine Turbidity Maxima with particular attention to the
upper Chesapeake Bay
-
Well-documented turbidity maxima are found all over the
world
Chesapeake Bay Hudson St. Lawrence Columbia San Francisco Bay
Chikugo Ems Gironde Weser Tamar ACE Basin
Seine Scheldt Jiaojiang Etc., etc.
-
Observations about particle trapping in ETMs
Particle trapping in ETMs occurs by asymmetrical tidal transport
of a pool of resuspendible particles with a limited range of
settling velocities
Fine sediments in estuarine environments almost always exist in
aggregated (flocculated) form. Aggregation and disaggregation can
be active processes, depending on concentration, stickiness, and
small scale shear.
Settling velocities of fine sediments trapped in ETMs are
determined by the aggregate properties (size and specific density),
not the individual particle properties
-
Infilling of the Old Susquehanna Channel
-
Upper Bay Bathymetry from a 3D charting program
-
Continuous, natural infilling of shipping channels requires
continuous maintenance dredging
-
Chesapeake ETM has been studied from physical and biological
perspectives in four
recent programs, 3 supported by NSF
Seasonal, short term studies during 5 of the last 7 years (3-9
days cruises in spring, summer, and fall) 1996 very wet 1998 wetter
than normal 1999 drought 2001 dryer than normal 2002 extreme
drought 2007 drought 2008 - ?
-
76.5 76.0
39.0
39.5
Baltimore
BWI
CBOS
Conowingo Dam
Axial CTD Surveys and Moored Axial CTD Surveys and Moored
DataData
-
10 20 30 40 50 60 70
Distance from Head of Bay [km]
-24
-20
-16
-12
-8
-4
0 TSS, (mg/l)
0
10
30
50
70
200
1000
Axial CTD Survey- May 2, 1996
10 20 30 40 50 60 70 80-24
-20
-16
-12
-8
-4
0F F F F SF SF SF SF E E E
Salinity, (PSU)
Dep
th (m
)D
epth
(m) 1 2 3
45
6
1 1
1
2 2 3
-12
-10
-8
-6
-4
-2
0
D
e
p
t
h
,
[
m
]
Salinity, [PSU]
-40
-40
-20
-20
-20
-200
0
020
20
20
40
40
40 60
6080
Time of Day, [hours]
-12
-10
-8
-6
-4
-2
0D
e
p
t
h
,
[
m
]
30
30
30 30
3030
30
30
30 30
30
40 40
40 40
40
40
40
40 40
50 50 50 50
6070
-12
-10
-8
-6
-4
-2
0
D
e
p
t
h
,
[
m
]
0
5
10
20
30
40
50
60
70
100
200
500
1000
ETM Channel Time Series- October 24 & 25, 1996
TSS, [mg/l]
Along Channel Current Speeds
-80-70-60-50-40-30-20-1001020304050607080
14 16 18 20 22 0 2 4 6 8 10 12 14 16
14 16 18 20 22 0 2 4 6 8 10 12 14 16
14 16 18 20 22 0 2 4 6 8 10 12 14 16
Physical Features of the ETMTSS= Total Suspended Solids
Ebb
Flood
-
Zooplankton in the ETM, May 1996
-
StripedBassLarvae
Spring 1996 TIES Program
500
0
70
0
No. m-3White PerchMorone americana
Striped BassMorone saxatilis
White PerchLarvae
-
Conceptual diagram of ETM sediment and zooplankton trapping at
the limit of salt
-
Annual Susquehanna River flows and suspended sediment loadings
to the upper Bay, 1991-1999
Monthly river flows and sediment loads to upper Bay during 1996,
compared to the average during the 1990s
-
Previous conclusion (Sanford et al. 2001): Particle trapping in
CB ETM highly dependent on settling speeds of flocculated fine
sediments, which vary seasonally
10 20 30 40 50 60 70 80-24
-16
-8
0
D
e
p
t
h
,
[
m
]
10 20 30 40 50 60 70 80Distance from Havre De Grace, [km]
-24
-16
-8
0
D
e
p
t
h
,
[
m
]
Salinity, [PSU]
TSS (mg/l)05102030405060701002005001000
SF SF SF SF SF SF SF E
Feb 1, 1996
Feb 1, 1996
10 20 30 40 50 60 70 80-24
-16
-8
0
D
e
p
t
h
,
[
m
]
10 20 30 40 50 60 70 80Distance from Havre De Grace, [km]
-24
-16
-8
0
D
e
p
t
h
,
[
m
]
Salinity, [PSU]
TSS (mg/l)05102030405060701002005001000
10 20 30 40 50 60 70 80-24
-16
-8
0
D
e
p
t
h
,
[
m
]
10 20 30 40 50 60 70 80Distance from Havre De Grace, [km]
-24
-16
-8
0
D
e
p
t
h
,
[
m
]
Salinity, [PSU]
TSS (mg/l)05102030405060701002005001000
SE F F F F F F F F F SF E E E E E E E SF F F F
Oct. 22, 1996
Oct. 22, 1996
Oct. 27, 1996
Oct. 27, 1996
February 1996 flood October 1996 flood
-
Near Bed Salinity Gradient v1 PSU
2001
10
20
30
40
50
60
10 20 30 40 50 60
1 PSU or dS/dx Location, [River Km]
E
T
M
L
o
c
a
t
i
o
n
,
[
R
i
v
e
r
K
m
]
1 PSU dS/dx 1:1 line
2002
10
20
30
40
50
60
10 20 30 40 50 60
1 PSU or dS/dx Location, [River Km]
E
T
M
L
o
c
a
t
i
o
n
,
[
R
i
v
e
r
K
m
]
1 PSU dS/dx 1:1 line
All Data
10
20
30
40
50
60
10 20 30 40 50 601PSU or dS/dx Location, [River km]
E
T
M
L
o
c
a
t
i
o
n
,
[
R
i
v
e
r
K
m
]
1PSU dS/dx 1:1 line
1996
10
20
30
40
50
60
10 20 30 40 50 60
1 PSU or dS/dx Location, [River Km]
E
T
M
L
o
c
a
t
i
o
n
,
[
R
i
v
e
r
K
m
]
1PSU dS/dx 1:1 line
-
Critical Stress and Erosion Rate Constant
0
0.1
0.2
0.3
0.4
0.5
0 0.1 0.2 0.3 0.4Critical stress, c (Pa)
E
r
o
d
e
d
m
a
s
s
(
k
g
m
-
2
)
y = -.0014+1.5688x1.6070
R2 = .78
0
0.1
0.2
0.3
0.4
0.5
0 0.001 0.002 0.003Erosion rate constant, M(kg s-1 Pa-1 m-2)
E
r
o
d
e
d
m
a
s
s
(
k
g
m
-
2
)
-
Spatial and Temporal Variation
0.0
0.1
0.2
0.3
May'01
July'01
Oct '01 May'02
July'02
Oct '02
E
r
o
d
a
b
l
e
m
a
s
s
(
k
g
m
-
2
)
Tolchester Grove Pt. TSL
-
Eroded Mass vs ws and Loading
0.0
0.1
0.2
0.3
May '01July '01 Oct '01 May '02July '02 Oct '02
E
r
o
d
a
b
l
e
m
a
s
s
(
k
g
m
-
2
)
0
2
4
6
8
S
e
t
t
l
i
n
g
S
p
e
e
d
(
m
m
s
-
1
)
Tolchester Grove Pt. Settling Speed
0
0.1
0.2
0.3
Mar'01
Jun'01
Sep'01
Dec'01
Mar'02
Jun'02
Sep'02
E
r
o
d
a
b
l
e
M
a
s
s
(
k
g
m
-
2
)
010203040506070
S
e
d
i
m
e
n
t
L
o
a
d
(
k
g
s
-
1
)
Tolchester GrovePt Sed load
-
Particle size and settling velocity measurementsModified
Valeportbottom withdrawal settling tube with water jacket and
reflective insulation
Video In-situ Settling Tube Apparatus (VISTA) mounted on
profiling rig with LISST, ADV, CTD; water pumped through tube,
valves closed, particle settling videotaped
Laser In-Situ Scattering and Transmission (LISST) particle size
in 32 bins between 2-500 microns, plus forward transmission
-
Disaggregated Sediment Sizes: high-volume filtering 1 mab
Sample Sampling Duration (h)Total Mass
(g)d25
(microns)d50
(microns)d75
(microns)
May Flood (25 cm s-1) 2.5 7.05
-
Video clip from May 2002 VISTA sample (approx 1 cm across)
-
10 20 30 40 50 60 70
20
10
0
D
e
p
t
h
(
m
)
D50 (microns)
25
50
75
100
150
200
250
300
10 20 30 40 50 60 70
20
10
0
Tot. Vol. Conc. (ul/l)
01020305075100150200300500100015002000
10 20 30 40 50 60 70
20
10
0
Vol.Conc. 66.5 um (ul/l)
01020305075100150200300500100015002000
10 20 30 40 50 60 70
20
10
0D
e
p
t
h
(
m
)
TSS (mg/l)
2
5
10
20
30
40
50
60
70
10 20 30 40 50 60 70River Km
20
10
0
D
e
p
t
h
(
m
)
Bulk Density (g/cm^3)
1.005
1.02
1.05
1.1
1.15
1.2
1.25
1.3
May 11, 2002 Strong Ebb Survey: LISST Data
-
10 20 30 40 50 60 70
20
10
0
D
e
p
t
h
(
m
)
D50 (microns)
25
50
75
100
150
200
250
300
10 20 30 40 50 60 70
20
10
0
Tot. Vol. Conc. (ul/l)
01020305075100150200300500100015002000
10 20 30 40 50 60 70
20
10
0
Vol.Conc. 66.5 um (ul/l)
01020305075100150200300500100015002000
10 20 30 40 50 60 70
20
10
0D
e
p
t
h
(
m
)
TSS (mg/l)
2
5
10
20
30
40
50
60
70
10 20 30 40 50 60 70River Km
20
10
0
D
e
p
t
h
(
m
)
Bulk Density (g/cm^3)
1.005
1.02
1.05
1.1
1.15
1.2
1.25
1.3
October 11, 2002 Weak Flood Survey
-
Tidal Cycle Anchor Station October 10, 2002: Flood - Ebb
8 9 10 11 12 13 14 15 16 17
10
5
0
D
e
p
t
h
(
m
)
0.10.511.522.533.544.5
8 9 10 11 12 13 14 15 16 17Time (EDT)
10
5
0
D
e
p
t
h
(
m
)
25
50
75
100
150
200
250
300
10
5
0
D
e
p
t
h
(
m
)
8 9 10 11 12 13 14 15 16 17
01020305075100150200300500100015002000
Shear (s-1) and current (cm s-1)
Total Volume (ul l-1) and TSS (mg l-1)
D50 (microns) and Salinity (psu)
-
Surface
y = 5.3443x1.0495
R2 = 0.45480.0001
0.001
0.010.00001 0.0001 0.001 0.01
d [m]
w
s
[
m
/
s
]
8:30 am9:30 am10:00 am10:30 am
Middle
y = 26.597x1.1912
R2 = 0.4544
0.0001
0.001
0.010.00001 0.0001 0.001 0.01
d [m]
w
s
[
m
/
s
]
8:50 am
10:20 am
Bottom
y = 1.908x0.8331
R2 = 0.47440.0001
0.001
0.010.00001 0.0001 0.001 0.01
d [m]
w
s
[
m
/
s
]
8:40 am
9:10 am
10:10 am
all depths
y = 147.15x1.3923
R2 = 0.60770.0001
0.001
0.010.00001 0.0001 0.001 0.01
d [m]
w
s
[
m
/
s
]
Surface
Middle
Bottom
Winterwerp 2002, Ems Estuary
LISST range
Owen tube
VISTA: VISTA: FlocFloc Settling Speed v. Settling Speed v.
FlocFloc Size, 10/10/2002Size, 10/10/2002
-
VISTA Results:VISTA Results:
depth ws,ave dave bulk D3 dr[m] [mm.s-1] [m] [g.cm-3] [-]
[m]
3 0.66* 170** 1.05 2.05 3.85
7 1.41* 243 1.05 2.19 3.15
9.5 2.26* 294 1.06 1.83 16.07
all 1.54 243 1.06 2.39 0.79
-
Valeport Settling Tube Results:
a)
0
2
4
6
8
10
0 50 100 150 200 250
TSS [mg . l-1]
w
s
5
0
[
m
m
.
s
-
1
]
b)
0
50
100
150
200
250
0 10 20 30 40 50
Non-settling [mg . l-1]
S
e
t
t
l
i
n
g
[
m
g
.
l
]
1996 (), 2001 (), 2002 ()
-
Conceptual Model of Particle Dynamics in Chesapeake Bay ETM:
-
Preliminary Conclusions Particle size usually smaller near
surface, but
evidence of large, watery flocs in pycnoclinesometimes d50
50-100 microns near surface, 100-200 microns
near bottom Only weak relationships to turbulent shear
Some very slowly settling particles always present, settling
particles appear to be resuspended and deposited tidally with a
broad range of settling speeds (ws50 0.4-8 mm/s) No relationship
between ws50 and concentration
apparent
-
www.BITMAXII.org
ETM (essentially) session at ERF 2007 (Houde and Sanford)
An Introduction to Sediment Transport in EstuariesOutlineGeneral
References for figures, etc. (not always noted)Basic
Characteristics of SedimentsGlobal Sediment Sources Rivers account
for 85% of inputs to global oceanEstuaries can trap large portions
of riverine inputs, can have large internal inputs from shoreline
erosion, and can import seThe bottom can serve as both source and
sink, and often dominates both terms in estuaries and coastal
seasBackground Classification of Sediment Transport (in decreasing
order of understanding)Now, what factors affect remobilization of
sediment from the bottom?Separation of total stress and skin
friction for quasi-steady flows (not the most sophisticated
technique, but has been used Surface gravity waves (only most
important aspects for sediment transport)Oscillatory boundary
layers and wave friction factorsSo surface gravity wave forcing can
play a dominant role in sediment transport in shallow, microtidal
environments (Nakagawa, Resistance to remobilization of bottom
sediments:Critical StressA threshold value of the applied skin
friction below which tErosion/Resuspension of silt/clay mixtures
(muds)Partial list of factors affecting fine sediment
erodibilityErodibility also can vary significantly in time and
space. For example, consolidation causes tc to increase rapidly
with deptErodibility can change significantly in response to
disturbance.Passage of a tropical storm, upper Chesapeake Bay, Sept
1992.Peter TraykovskiHigh Resolution Acoustic Backscatter Profiler,
Hudson ETMRecent emphasis in field has been on development of new
techniques for site-specific, in-situ erosion testingIn situ
annular and linear flumesIn situ small scale erosion chambersQuasi
in situ core testingHuge scatter in reported erosion rates may be
due to several factors:How do experimental design and data analysis
procedures affect interpretation of erosion data? For illustration,
consider datWe have known for years that a rapid increase of tc
with depth into the bed results in a rapidly time varying erosion
rate:If erosion behavior is time dependent, then differences in the
time history of stress application affect the results of
erosioDifferences in interpretation of critical stress are also
associated with time/depth dependenceSea Carousel v. Microcosm
comparison using individual data analysis and interpretation
techniquesTime dependent erosion behavior is relatively
straightforward to derive theoretically, especially for specific
erosion formulaReanalysis of Sea Carousel data using the Microcosm
approach reveals real similarities and differencesMeasurement
ConclusionsMeasurement Conclusions, continuedFrom Sanford and Halka
(1993)Modeling Cohesive Sediments: Erosion, Deposition, and Bed
ProcessesWhat are the main challenges for fine sediment transport
research and application?Bottom Sediment
Erosion/ResuspensionBiological effectsConcentrated benthic
suspensionsSpatial heterogeneityModelingNOPP CSTM participants and
their rolesSelected cohesive sediment erosion
formulationsConsolidation effects on erosion Time dependent erosion
behavior is straightforward to derive for specific erosion
formulations :What would an ideal cohesive bed/erosion/deposition
model include?Selected Existing Cohesive Sediment Transport
ModelsSediment Bed Model Development in ROMSSED based on Sanford
(in press)How well does an exponential approach to equilibrium
approximate consolidation?Comparison to data of Toorman and
Berlamont (1Example: Erosion, deposition, and consolidation of a
pure mud and a sand-mud mixtureModeling ConclusionsModeling
Conclusions, continuedEstuarine Turbidity Maxima with particular
attention to the upper Chesapeake BayWell-documented turbidity
maxima are found all over the worldObservations about particle
trapping in ETMsInfilling of the Old Susquehanna ChannelContinuous,
natural infilling of shipping channels requires continuous
maintenance dredgingChesapeake ETM has been studied from physical
and biological perspectives in four recent programs, 3 supported by
NSFZooplankton in the ETM, May 1996Previous conclusion (Sanford et
al. 2001): Particle trapping in CB ETM highly dependent on settling
speeds of flocculated fineNear Bed Salinity Gradient v1 PSU
Critical Stress and Erosion Rate ConstantSpatial and Temporal
VariationEroded Mass vs ws and LoadingParticle size and settling
velocity measurementsDisaggregated Sediment Sizes: high-volume
filtering 1 mabVideo clip from May 2002 VISTA sample (approx 1 cm
across)May 11, 2002 Strong Ebb Survey: LISST DataOctober 11, 2002
Weak Flood SurveyTidal Cycle Anchor Station October 10, 2002: Flood
- EbbValeport Settling Tube Results:Conceptual Model of Particle
Dynamics in Chesapeake Bay ETM:Preliminary Conclusions
